基于YOLOv5的轻量级中草药识别研究

Chinese Herbal Medicine Recognition Based on Improved YOLOv5 Algorithm

中草药是中医药学的重要组成部分。随着中医药国际地位的提高,中草药的使用更加广泛。然而,传统的人工识别中草药方法效率低且易出错。为解决这一问题,提出一种基于YOLOv5目标分类检测算法的中草药识别方法。对YOLOv5神经网络结构进行改进,提高特征表示的质量和判别力。采用迁移学习策略,提升对小目标和小样本类别的检测精度。研究可应用于中药材溯源、质量控制等领域,为中医药现代化和国际化提供技术支撑。

Chinese herbal medicine is an important part of traditional Chinese medicine.With the improvement of the international status of Chinese medicine,the use of Chinese herbal medicine is more extensive.However,traditional methods of manual identification of Chinese herbs are inefficient and error-prone.To solve this problem,this paper proposes a method of Chinese herbal medicine recognition based on YOLOv5 target classification detection algorithm.The structure of YOLOv5 neural network is improved to improve the quality and discriminability of feature representation.Transfer learning strategy is adopted to improve the detection accuracy of small targets and small samples.The research can be applied to the traceability of Chinese medicinal materials,quality control and other fields to provide technical support for the modernization and internationalization of Chinese medicine.